SECARB Phase III Citronelle Project (Anthropogenic Test) in Alabama

Transcription

1 SECARB Phase III Citronelle Project (Anthropogenic Test) in Alabama Dave Riestenberg, PM, ARI Rob Trautz, SR. TECH MGR., EPRI George Koperna, VP, ARI Steve Carpenter, VP, ARI 10th Annual SSEB Stakeholders Meeting 12 March 2015

2 Acknowledgement This presentation is based upon work supported by the Department of Energy National Energy Technology Laboratory under DE-FC26-05NT42590 and was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

3 Topics of Discussion 1. BRIEF Introduction 2. Monitoring & Modeling Lines of Evidence 3. AoR & Modeling Update 4. Lessons Learned 5. Questions, Answers, Discussion 3

4 Project Objectives 4 1. Support the United States largest prototype coal-fired CO 2 capture and transportation demonstration with injection, monitoring and storage activities; 2. Test the CO 2 flow, trapping and storage mechanisms of the Paluxy; 3. Demonstrate how a saline reservoir s architecture can be used to maximize CO 2 storage and minimize the areal extent of the CO 2 plume; 4. Test the adaptation of commercially available oil field tools and techniques for monitoring CO 2 storage 5. Test experimental CO 2 monitoring activities, where such technologies hold promise for future commercialization; 6. Begin to understand the coordination required to successfully integrate all four components (capture, transport, injection and monitoring) of the project; and 7. Document the permitting process for all aspects of a CCS project.

5 Storage Site: The Citronelle Oilfield 5

6 UIC Class V Permit Class V Experimental Injection Well permit Short duration of injection (3 years) and modest volumes of CO 2 Characterization and modeling of stacked CO 2 storage CO 2 injection under real world operating conditions Demonstration of experimental monitoring tools and methods Many Class VI (CO 2 sequestration well) standards were applied Injection Area of Review (AOR) determined by modeling and monitoring results (updated annually) Deep, shallow and surface CO 2 monitoring Injection stream monitoring Site closure based on USDW non-endangerment demonstration and CO 2 containment (5-yr renewal) Based on monitoring and modeling results 6

7 Storage Project Status ADEM granted permission to inject on August 8, 2012 Injection commenced on August 20, 2012 Through September 1, 2014, approximately 114 thousand metric tons of CO 2 were injected A crosswell seismic survey acquired in June, 2014 captured a time-lapse image of the CO 2 plume Other testing and monitoring activities indicate containment The project entered the Post-Injection Site Care Period in September, 2014 Site closure based on CO 2 containment and USDW nonendangerment 7

8 CO 2 Injection History 8

9 1. Monitoring & Modeling Lines of Evidence 9

10 Anthro Test MVA Program Drivers Multiple lines of evidence to confirm CO 2 containment Soil CO 2 flux and tracer monitoring Crosswell seismic and VSP Pulsed neutron capture logging (in- and above-zone saturation) USDW monitoring Assure non-endangerment of USDWs Monitoring geochemistry of shallow aquifers Test experimental methodologies Inform the reservoir simulation Above tools plus downhole pressure 10

11 MVA Elements and Frequency 11 MVA Method Shallow Soil flux Groundwater sampling (USDW) PFT survey Deep CO2 volume, pressure & composition Reservoir fluid sampling Injection, temperature & spinner logs Pulse neutron logs Crosswell seismic Vertical seismic profile (VSP) Experimental Distributed Temperature Sensing (DTS) Comparative fluid sampling methods MBM VSP Distributed Acoustic Sensing (DAS) MBM VSP & OVSP Seismic Frequency Continuous Monthly Quarterly Annual Milestone (Baseline, Injection, Post)

12 Anthro Test MVA Program One new injector (D-9-7 #2) Two new deep observation wells (D-9-8 #2 & D-9-9 #2) Two in-zone & above zone monitoring wells (Citronelle wells D-4-13 & D-4-14) One PNC logging well (D-9-11) 12 soil flux monitoring locations PFT monitoring on nine well pads Crosswell and VSP between D-97#2 and D-9-8#2 12

13 CO 2 Containment- Soil CO 2 Flux and Tracer Monitoring Soil CO 2 Flux Tracer Results Inoculation Testing Well/Sample AUG 2012 JUN 2013 NOV 2013 D-9-1 ND ND ND D-9-2 ND ND ND D-9-3 ND ND ND D-9-6 ND ND ND D ND ND ND D-9-8 Invalid Data ND ND D-9-9 ND ND ND D-9-10 Invalid Data ND ND D-9-11 ND ND ND Air Blank 1 ND NST NST System Blank ND ND Soil CO 2 results appear to vary as a function of mean temperature and PFT have been non-detect 13

14 Deep Monitoring Time-Lapse Crosswell Seismic Crosswell seismic surveys allow for high-resolution mapping of the acoustic travel time (velocity) and seismic reflectors between a pair of wells When CO 2 displaces water in the formation, it changes the acoustic impedance of the rock Acoustic wave decreases and its direct travel time increases Results from repeat surveys performed during or after CO 2 injection can be compared to a pre-injection baseline survey to image the extent of the CO 2 plume (referred to as time-lapse imaging ) Baseline and repeat 2-D crosswell seismic surveys were performed between the injection well and the observation well 14

15 Crosswell Survey Configuration and Parameters Pre-injection baseline survey acquired on January 19-26, 2012 Repeat survey was acquired on June 14-23, 2014 Source Type: Piezoelectric deployed in D-9-7#2 well Receiver type: Hydrophone 10 levels deployed in D-9-8#2 well 840 between D-9-7#2 and D-9-8#2 at reservoir depth Schematic showing the open well completion in observation well D-9-8 during the baseline survey (left) and packer/tubing completion during the repeat (right) Receivers were deployed in the open well during the baseline survey and inside the MBM tubing/packer assembly during the repeat survey, thus changing the baseline conditions 15

16 Injection Zone Baseline Survey Results Velocity tomography and reflection imaging (right) provided a good representation of the reservoir and confining unit ~10 feet vertical resolution No reservoir or confining unit discontinuities or small-scale faults were observed in the reflection data Layering observed in the Upper Paluxy will help disperse the CO 2 plume, thus minimizing its footprint Baseline velocity tomogram should be of sufficient quality for timelapse CO 2 plume imaging Confining Zone Composite image mapping the seismic reflections (squiggles) superimposed on top of the velocity tomogram (colored background) 16

17 Comparison of Baseline and Repeat Data Quality First arrivals and reflection data from the baseline survey have strong amplitudes and little noise, representing good quality data The first arrivals for the repeat survey are fairly weak probably due to signal attenuation caused by deploying the hydrophones inside the stiff production tubing and packer First Arrivals or Picks Poor quality reflection data The reflection data that follow the first arrivals are noisy and of poor quality for the repeat survey Side-by-side comparison of a baseline (left) and repeat (right) shot gather There is a noticeable decrease in the signal-to-noise ratio (SNR) between the baseline and repeat surveys, which limits data interpretation 17

18 Injection Zone Comparison of Crosswell Reflectors Baseline Tomogram Strong, continuous reflectors Repeat Tomogram Weak and/or discontinuous reflectors Confining Zone No reflector was detected at or near the top of the CO 2 where one should be present Reflection data from the repeat survey are of poor quality and limited use. Likely cause is interference by tube waves moving up and down the well 18

19 Time-Lapse Differencing Using the Baseline and Repeat Velocity Tomograms Injection Zone First arrivals from repeat survey were of sufficient quality to produce a velocity difference image (right) showing regions where seismic velocity has changed over time Time-lapse difference image indicates a decrease in seismic velocity in the upper injection zone of up to 3%, suggesting an increase in CO 2 saturation No significant negative velocity anomalies Decrease in velocity (negative anomaly) Confining Zone More importantly, no negative velocity anomalies are observed in or above the confining unit implying no detectable leakage out of inj. zone Pixelized difference tomography results without seismic reflection overlay showing positive velocity differences in warm colors and negative differences in cool colors 19

20 Plume Image Comparison with Spinner Surveys J I H G F E D C B A Time-lapse image shows CO2 plume located primarily in Paluxy sands F-H October 2013 spinner survey show these sands taking only 10% of the flow Sand Sand Unit Properties (ft) Nov 2012 Aug 2013 Oct 2013 Unit Bottom Top Thickness Flow % Flow % Flow % J 9,454 9, I 9,474 9, H 9,524 9, G 9,546 9, F 9,580 9, E 9,622 9, D 9,629 9, C 9,718 9, B 9,744 9, A 9,800 9,

21 Deep MVA Pressure Response D9-8#2 D4-14 In Zone D4-13 Above Confinement CO 2 Injected Downhole pressure data is a primary input to the history match and plume model 21

22 RPI CO 2 Injection Production Analysis In the plot below, the slope of each line is related to the transmissibility (kh) of the reservoir and the y-intercept is related to the wellbore-reservoir connectivity 10 RPI P WeightRate 5 0 Slope is a function of kh Elevation on Y axis is a function of connectivity to wellbore (i.e. x f or skin) Slope is unchanged kh has not changed Increased elevation on Y axis indicates damage to wellbore connectivity Slope is has increased kh has dropped No change in elevation on Y axis indicates wellbore connectivity unchanged 0 10 Log of Time 22

23 Tubing Pressure Daily CO 2 Rate, Metric Tons D-9-7#2 Injection Pressure History 3,000 2,500 Transient # Transient # , , , June, Repeat Crosswell Injection well killed, tubing pulled Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Oct-13 Jan-14 Apr-14 Jul-14 Nov-14 spinner survey Notable Increase in injection tubing pressure after crosswell seismic workover 23

24 CO 2 Injection Production Analysis Post-Workover Transient #1 Pre-Workover Decrease in connectivity and transmissibility (kh) after D-9-7#2 workover 24

25 Deep MVA Pressure Response Time The system, as expected, is getting more compressible with continued injection. As a result, the pressure transient travel time between the injection and observation wells continues to grow. D-4-14 (in zone) D-9-8#2 25

26 Citronelle Groundwater Sampling Program Three dedicated groundwater sampling wells and one water well Well Depth (ft) Elev. (ft) D9-9 MW D9-7 MW-2S D9-7 MW-2D D9-8 WW Three background sampling events prior to CO 2 injection Nine quarterly sampling events since injection started 17 metals, alkalinity, TDS, TIC, ph etc. Groundwater sampling locations (circled) 26

27 Metals, ph, inorganic carbon, etc Occur Naturally in Groundwater Complicating Regulatory Interpretation UIC Permit Backgrd Backgrd Backgrd Backgrd Analyte Unit Level MW-1 MW-2S MW-2D WW Metals, Total Aluminum ug/l 200 < <100 < < Antimony ug/l 6 <5 <5 <5 <5 Arsenic ug/l 10 <5 <5 <5 <5 Barium ug/l < Berylium ug/l 4 <3 <3 <3 <3 Cadmium ug/l 5 <5 <5 <5 <5 Chromium ug/l 100 <5 <5 <5 13 <5 Copper ug/l 1,300 <10 <10 <10 < Iron ug/l < Lead ug/l 15 <5 <5 <5 5.5 <5 35 Manganese ug/l < Mercury ug/l 2 <0.2 <0.2 <0.2 <0.2 Nickel ug/l 100 <5 <5 <5 5.1 <5 Selenium ug/l 50 <10 <10 <10 <10 Silver ug/l 100 <5 <5 <5 <5 Thallium ug/l 2 <1 <10 <1 <1 <1 Zinc ug/l 5000 <20 25 <20 <20 69 <20 21 General Chemistry Alkalinity, Bicarbonate as CaCO3 mg/l Monitor Alkalinity, Total mg/l Total Dissolved Solids mg/l Total Inorganic Carbon mg/l Monitor Field Parameters ph Std units

28 Citronelle Compliance Monitoring Program is Based on U.S. Environmental Protection Agency RCRA Guidelines Compliance data exceed UIC permit levels? Yes Compliance data exceed natural background range? Yes Compliance data exceed Upper or Lower Confidence Limit (C.L.)? No - Check for outliers - Check for normality - Compute 95% C.L. - Value to value comparison with C.L. Continue Yes Monitoring Wilcoxon-Mann-Whitney Trend Analysis. Trend Observed? Yes Take action 28

29 Summary of Intra-Well Statistics Intra-well statistical analyses indicates several significant trends for the N=3 through 12 sampling events ph decrease at D-9-7 MW-2D Ba, Fe, Mn and sulfate increase at D-9-7 MW-2S Ba, Fe and Mn increase at D-9-9 MW-1 The statistical analysis would imply that the groundwater quality is changing at Citronelle, but is CO 2 the cause? 29

30 Time Series Trends (N=1 through 12 Sampling Events) 30

31 Multiple Lines of Evidence are Needed to Prevent False- Negatives and Positives 1. Multiple lines of evidence for the potential influence of carbon dioxide at individual monitoring wells have not been identified to date at Citronelle 2. Groundwater systems are inherently complex requiring thoughtful design of the background sampling and compliance monitoring programs 31

32 2. AoR & Modeling Update 32

33 Introduction CMG s GEM software was used to model the injection into the upper Paluxy Sandstone and forecast the subsurface movement and pressure profile of the CO 2 in order to meet our Class V UIC AoR guidelines. Modeling was done in a three-step process. These steps were: History matching the injection through 31 AUG Forecasting continued injection through the end of the proposed injection period to account for anticipated interruptions (reserve shut-down from 01 NOV APR 2014). Then, the plume was allowed to relax to understand pressure equilibration and plume stabilization. 33

34 Injection Profile Match Field gathered injection profile surveys were used to understand how the CO 2 entered the Paluxy reservoir sand bodies (8 sands, 10 sets of perforations). Simulation

35 Pressure, psia Offset Pressure Profile Match Pressure data was used to calibrate modeled upper Paluxy sand layer (perf sets 1 and 2) pressure responses at 870 ft (265m) at D 9-8 #2 and more than 3,000 ft (914m) at the D Note: the late time erratic pressure data in the D 9-8 #2. Upper Paluxy Pressure Match at the D 9-8 #2 Upper Paluxy Pressure Match at the D 4-14 SECU D4-14 Pressure Aug-12 Sep-12 Nov-12 Dec-12 Feb-13 Apr-13 May-13 Jul-13 Sep-13 Oct-13 Date D4-14 Bottom D4-14 Top D4-14 Simulation 35

36 Predicted Radial CO 2 Plume Extent on August 31, 2013 of 440 ft (134 m) As expected, the two most porous and permeable sands accepted the bulk of the CO 2. 36

37 Predicted Radial CO 2 Plume Extent on October 31, 2014 of 720 ft (219 m) Continuing the injection (135,000 tonnes more), the plume continues to spread across the 9460 and 9620 sands. 37

38 Predicted Radial CO 2 Plume Extent on October 31, 2017 of 720 ft (219 m) Despite the high permeability of the formation and a gentle dip, there is not a substantial change in the plume extent in the three years following the cessation of injection. In fact, the plume extent has equilibrated. 38

39 3. Lessons Learned 39

40 Lessons Learned Time and cost reductions realized, but not yet commercial Data, data, & more data MVA systems can impact injection and vice versa We have good reservoir capacity and injectivity apparent injection damage after workover 40

41 Lessons Learned Data resolution may be challenging in deep settings Every potential storage project is different & MVA should be site specific in design When deploying non-commercial MVA protocols, redundancy with more commercial tools is necessary to ensure the data quality Build from the lessons learned at existing projects 41

42 4. Questions, Answers, Discussion 42

43 Together Shaping the Future of Electricity 43

44 EXTRAS 44

45 USDW Protection - Groundwater Geochemistry 3 - Background Monitoring Events: January 2012 (N=1) through July 2012 (N=3) 10 - Injection Period Monitoring Events: November 2012 (N=4) through February 2015 (N=13) Background anomalies of Mn, Fe, and Cl above UIC permit discharge limits. To evaluate the potential exceedance of regulatory standard (e.g., UIC permit discharge limit), the EPA GW Unified Guidance recommends statistical comparisons ( value to value comparison to standard and evaluation of changes between baseline and monitoring) Quarterly testing to continue throughout the PISC 45

46 USDW Monitoring Well Network Monitoring Well D-9-7 MW-2S D-9-7 MW-2D D-9-9 MW-1 Well Depth Well TOC Elevation Screened USDW (ft. BTOC) (ft. AMSL) Miocene-Pliocene Aquifer Miocene-Pliocene Aquifer Miocene-Pliocene Aquifer Water Supply Well Near D-9-8 Miocene-Pliocene Aquifer/Watercourse Aquifer ~143 Not Surveyed 46

47 Comparison of Baseline Values to Permit Analyte D-9-7 MW2D UIC Permit Discharge Limits (µg/l) Range of Valid Background Concentrations (µg/l) Aluminum 200 < Antimony 6 <5 Arsenic 10 <5 Barium 2,000 <10-29 Beryllium 4 <3 Cadmium 5 <5 Chromium 100 <5-13 Copper 1,300 <10 Iron 300 < Lead 15 <5-5.5 Manganese 50 <10-18 Mercury 2 <0.2 Nickel 100 <5-5.1 Selenium 50 <10 Silver 100 <5 Thallium 2 <1 Zinc 5,000 <20-69 Range of values because of the small background data set (N=3). Selected naturally occurring background concentrations which exceed UIC Permit discharge. 47

48 Statistically Determined Potential Lines of Evidence for CO 2 Influence* Multiple lines of evidence for the potential influence of CO 2 at individual monitoring wells have NOT been identified *e.g. Wilkin and Digiulio (2010) 48